Download Anxious and Depressive Disorders and Their Comorbidity

Anxious and Depressive Disorders and Their
Comorbidity: Effect on Central Nervous System
Noradrenergic Function
Oliver G. Cameron, James L. Abelson, and Elizabeth A. Young
Background: Although comorbidity of anxiety with depression is common, investigations of physiologic abnormalities related
specifically to comorbidity are rare. This study examined relationships of DSM-IV-defined depression, anxiety, and their comorbidity
to noradrenergic function measured by blunting of the growth hormone (GH) response to the alpha2 adrenoreceptor agonist (and
imidazoline receptor agent) clonidine and by blood pressure and symptom responses.
Methods: Fifteen subjects with pure social anxiety or panic disorder, 15 with pure major depression, and 18 with both depression and
anxiety were compared with healthy control subjects matched for age and gender. Other factors known to affect GH (weight, menstrual
status, prior antidepressant, or other drug exposure) were controlled.
Results: Anxiety produced GH blunting, but depression was associated with normal GH responses. The comorbid state did not affect
results beyond the impact of anxiety. Preclonidine stress-related GH elevations were observed, to the greatest degree in anxious subjects.
Relevant symptom, but not blood pressure, changes were significantly associated with blunting.
Conclusions: With use of pure depression and anxiety groups and careful control of other factors known to affect GH, these results
demonstrate central nervous system noradrenergic dysfunction in anxiety disorders. In contrast to less rigorously controlled studies,
noradrenergic function in depression was normal.
Key Words: Anxiety, clonidine, comorbidity, depression, growth
hormone, noradrenergic
C
o-occurrence of psychiatric disorders (comorbidity) is
common. In persons with depression or an anxiety disorder, comorbidity with the other disorder occurs in
one-quarter to one-half of individuals (Kessler et al 1994, 1998,
1999) and is associated with increased severity (Barbee 1998;
Lydiard 1991; Roy-Byrne et al 2000). Although there has been
extensive interest in biological factors associated with depression
and anxiety disorders, biological changes associated with the
comorbid state, defined by diagnoses or symptoms, have been
neglected. Individuals with comorbid diagnoses have either been
excluded or have been treated as part of the group that has the
same primary (occurring first) or predominant (more severe)
diagnosis as the comorbid individuals.
One biological system of major interest in depressed and
anxious individuals is brain noradrenergic function. One marker
of this function is the growth hormone (GH) increase produced
by stimulation with the alpha2 adrenoreceptor agonist clonidine.
Investigations of this marker in depression have reported blunted
GH responses, implying reduced noradrenergic function, responsivity, or both (e.g., Amsterdam and Maislin 1990; Ansseau
et al 1988; Charney et al 1982b; Checkley et al 1981; Correa et al
2001; Horton et al 1986; Siever and Uhde 1984), including studies
indicating blunting is a trait marker (Charney et al 1982a; Mitchell
et al 1988; Siever et al 1992); however, a number of studies found
no blunting (Fu et al 2001; Gann et al 1995; Katona et al 1993;
Mitchell et al 1991; Schittecatte et al 1989).
Single anxiety studies have reported blunting in generalized
anxiety (Abelson et al 1991) and posttraumatic stress disorder
From the Department of Psychiatry (OGC, JLA, EAY) and Mental Health
Research Institute (EAY), University of Michigan Medical Center, Ann
Arbor, Michigan.
Address reprint requests to Oliver G. Cameron, M.D., Ph.D., 1215 Southwood
Court, Ann Arbor, MI 48103-9735; E-mail: [email protected].
Received January 15, 2004; revised July 7, 2004; accepted August 19, 2004.
0006-3223/04/$30.00
doi:10.1016/j.biopsych.2004.08.007
(Morris et al 2004). Blunting occurred in most (e.g., Abelson et al
1992; Brambilla et al 1995; Charney et al 1992) but not all
(Schittecatte et al 1992) panic studies, including one suggesting
blunting is a trait marker (Coplan et al 1995). Limited results in
social phobia (Tancer et al 1993, 1994/5) and obsessive– compulsive disorder (Hollander et al 1991; Lee et al 1990; Siever et al
1983) were inconsistent (blunted or normal).
Without drug administration, GH was abnormal during phobic exposure (Curtis et al 1979) and during “spontaneous”
(Cameron et al 1987) and situationally induced (Woods et al
1987) panic attacks. It was also abnormal in response to isoproterenol-induced symptoms (Nesse et al 1984). In depression, GH
abnormalities have been associated with other pharmacologic
challenges (Ansseau et al 1988; Hasey et al 1985; Matussek and
Laakmann 1981) and under baseline conditions without drug
(Mendlewicz et al 1985). Several factors (gender, age, weight,
menstrual status and cycle phase, and prior antidepressant or
other drug exposure) affect the GH response to clonidine
(Altomonte et al 1988; Gil-Ad et al 1984; Kimber et al 2001;
Tancer et al 1990).
Clonidine also binds to imidazoline receptors (Bousquet et al
1998; Regunathan and Reis 1996). Growth hormone release can
result from either alpha2 adrenoreceptors or imidazoline receptor stimulation (Balldin et al 1993; Bamberger et al 1995).
Sedative effects of clonidine seemingly are mediated by noradrenergic receptors, and the hypotensive effects by imidazoline
receptors (Fornai et al 1990; Schafer et al 1995). Thus, these
markers might be used to differentiate these receptor effects.
Blunted GH responses to clonidine in anxiety disorders, high
comorbidity between depression and anxiety, and studies suggesting that, with careful experimental control, depressed individuals might be normal on this measure, all imply that past
positive results in depression could be due to comorbid anxiety.
Furthermore, the peripheral (Balon et al 1990; Cameron and
Minoshima 2002; Nesse et al 1984; Shear 1986) or central nervous
system (CNS; Cameron et al 2000; Charney et al 1984; Gurguis et
al 1997) adrenergic activation producing anxietylike symptoms,
similarity of some naturally occurring anxiety symptoms to
adrenergic activation (Cameron 1994), reduction of such sympBIOL PSYCHIATRY 2004;56:875– 883
© 2004 Society of Biological Psychiatry
876 BIOL PSYCHIATRY 2004;56:875– 883
toms by adrenergic blockade (Lader 1988; Tyrer 1988), and the
intimate involvement of CNS noradrenergic systems including
the locus ceruleus in fear (Bremner et al 1996; Sullivan et al 1999;
Tanaka et al 2000)—with none of these factors directly involved
in depression—all suggest that adrenergic functions including
that measured by GH responses to clonidine are directly related
to anxiety but not depression. To test this hypothesis, individuals
comorbid for major depression and an anxiety disorder were
studied, along with individuals with pure major depression or a
pure anxiety disorder. All were compared to specifically matched
control subjects. Based on possible confoundings recognized in
the GH depression literature, including comorbidity, as well as
research documenting involvement of noradrenergic systems in
anxiety (Cameron and Nesse 1988; Johnson and Lydiard 1995;
Ressler and Nemeroff 2000; Sullivan et al 1999) and anxietyassociated functions (e.g., apprehension, arousal, attention, novelty, stress; Aston-Jones et al 1999; Koob 1999; Robbins 1984;
Stanford 1995), we hypothesized that abnormal brain noradrenergic function, as indicated by GH blunting to clonidine, implying diminished postsynaptic alpha2 adrenoreceptor effects,
would be associated with anxiety but not depression. What roles
symptom severity and the presence of comorbidity per se play in
GH blunting were also investigated.
Methods and Materials
Subjects
Ninety-six subjects were recruited by advertisement, including 15 with an anxiety disorder without depression (anxiety
group), 15 with major depression without an anxiety disorder
(depression group), 18 with both disorders (comorbid group),
and 48 individually age and gender matched control subjects.
Diagnoses were made according to DSM-IV criteria with a
Structured Clinical Interview (SCID) interview by an experienced
research nurse. Subjects with pure disorders did not reach
present or past criteria within 2 years for the other disorder, and
the pure disorder was always primary, whereas comorbid individuals either met criteria for both disorders currently or within
the past year. Control subjects never had any psychiatric disorders themselves or in any first-degree relatives. Subjects with
pure or comorbid anxiety had predominant anxiety diagnoses of
social phobia or panic disorder. Most were naive to psychotropic
medications; five had received an selective serotonin reuptake
inhibitor. All were medication free for at least 9 months. Obese
subjects (body mass index ⱖ 30) or those with any eating
disorder or a recent weight loss averaging more than 1 kg/week
were excluded, as was anyone under 18 or over 50. Female
subjects were premenopausal and were studied within 10 days of
the onset of menstruation. The University of Michigan Medical
Center IRB approved the study. All subjects gave written informed consent.
Procedures
Subjects were admitted to the Clinical Research Center at 7:30
AM. They were studied at bed rest with an intravenous catheter
for clonidine infusion and blood sampling. Automated measurement of blood pressure (BP), blood sampling, and completion of
self-rated visual analogue scales (VAS; “drowsy” rating) were
performed before clonidine administration, and after 15, 30, 60,
90, and 120 min. Clonidine (Duraclon R), 2.0 ␮g/kg, was
administered intravenously over 5 min by infusion pump, starting at 9:00AM. Preclonidine measurements were done for GH at
30, 15, and 1 min before clonidine, for BP at 60 and 10 min
www.elsevier.com/locate/biopsych
O.G. Cameron et al
before, and for VAS ratings at 10 min before. The State and Trait
Anxiety Inventories (STAI) were completed at 8:50 AM and the
Hamilton Depression and Anxiety Rating Scales and the Sheehan
Disability Scale between 9:30 and 10:30 AM. Separated by at least
2 days before or after this study, each subject had a Trier Social
Stress Test (TSST; Kirschbaum et al 1993), involving performance
of a 5-min speech and a difficult mathematical subtraction task
with time pressure in front of an audience of three “experts”
(procedure detailed in Young et al 2004). Analyzed also were the
STAI State and three VAS ratings (“nervous,” “anxious,” “fearful”)
obtained during the TSST immediately before the social challenge. Order did not affect results of either study.
Blood samples for GH (ng/mL) and somatomedin-C (ng/
mL) were stored on ice for a maximum of 30 min, then plasma
was separated and frozen at –70°C. Growth hormone was
assayed by Nichols IRMA method (interassay coefficient of
variation of 6.8%). Somatomedin-C was assayed locally by
radioimmunoassay.
Data Analyses
Area under the curve (AUC) was calculated by trapezoidal
approximation for relevant data points. Distributions of untransformed data were approximately normal and log-transformation
did not visibly affect normality, so raw data were used in all
analyses except AUC. In addition to repeated-measures data,
patients were compared with control subjects on GH peak
change scores, calculated as highest GH level postclonidine
(always either 9:30 or 10:00 AM) minus last GH level before
clonidine administration (9:00 AM).
Hypotheses were tested using t tests and analyses of variance
(ANOVA), including repeated measures (RM-ANOVA) with covariance (RM-ANCOVA). Because GH levels were measured both
before and after clonidine and because GH measurements after
clonidine were timed to start before a robust response and to end
after the response was complete, interaction tests of each RMANOVA and RM-ANCOVA, testing for nonparallel levels between
groups over time (e.g., a blunted GH clonidine response), were
used as the main hypothesis tests. Because age and gender can
affect GH levels and patient groups could not be closely matched
on these variables, and because assay variability over time can
affect results, analyses always compared patients with their
individually matched control subjects, with subject and matched
control always assayed in the same batch. Regressions were used
to assess relationships of preclonidine to postclonidine GH
levels, and relationships of age, weight, and symptom ratings to
GH levels.
First, the hypothesis that patients (regardless of diagnosis)
differed from control subjects was examined by comparing all
patients combined with all control subjects. Then, to dissect the
relative contributions of anxiety, depression, and the comorbid
state to any GH blunting, two approaches were used. First, each
of the DSM-IV-defined pure anxiety, pure depressed, and comorbid groups were compared with their respective control subjects.
The comorbid patients were then subdivided into two groups,
based on predominant symptoms, to identify the disorder more
severely affecting the subject at time of study. All comorbid
subjects except one could be assigned a predominant diagnosis,
based on the more symptomatic or dysfunction-producing diagnosis, for the DSM-IV-defined qualifying duration of symptoms—
for example, 2 weeks for a major depressive episode or 1 month
for panic disorder. All patients with predominant anxiety (n ⫽
23) were compared with their control subjects and all with
predominant depression (n ⫽ 24) to theirs.
BIOL PSYCHIATRY 2004;56:875– 883 877
O.G. Cameron et al
Table 1. Subject Characteristics of the Pure Depression, Pure Anxiety, Comorbid, and Control Groups (mean ⫾ SD, where applicable)
n
Age (years)
Gender (F:M)
Weight (kg)
Anxiety Diagnoses
Panic disorder
Social phobia
Depression Diagnoses
Melancholic
Nonmelancholic
Hamilton Depression Ratinga
Hamilton Anxiety Ratingb
STAI-Traitc
STAI-Stated
TSST STAI-Statee
Sheehan Scalef
Depression
Comorbid
Anxiety
All Control
Subjects
15
28.7 ⫾ 7.7
6:9
70.9 ⫾ 13.4
0
0
0
15
7
8
19 ⫾ 4.0
13.8 ⫾ 5.5
50.4 ⫾ 11.2
42.0 ⫾ 14.1
35.4 ⫾ 9.3
17.9 ⫾ 5.7
18
23.2 ⫾ 5.5
13:5
67.1 ⫾ 7.6
18
5
13
18
4
14
12.7 ⫾ 6.7
12.4 ⫾ 7.0
53.3 ⫾ 8.3
50.3 ⫾ 12.1
45.6 ⫾ 9.4
20.6 ⫾ 7.0
15
25.4 ⫾ 8.5
8:7
70.3 ⫾ 13.9
15
3
12
0
0
0
3.3 ⫾ 3.4
5.8 ⫾ 7.0
40.1 ⫾ 11.0
34.1 ⫾ 7.4
43.2 ⫾ 11.1
9.0 ⫾ 6.5
48
26.0 ⫾ 7.1
27:21
68.7 ⫾ 12.1
0
0
0
0
0
0
.5 ⫾ .9
.7 ⫾ .7
27.0 ⫾ 5.2
25.0 ⫾ 4.3
28.6 ⫾ 8.6
1.4 ⫾ 1.1
STAI, State and Trait Anxiety Inventory; TSST, Trier Social Stress Test.
Analysis of variance (ANOVA): F ⫽ 137.9, p ⬍ .0001 for group; by Fisher protected least significant difference (PLSD), p ⬍ .05 for all group contrasts.
ANOVA: F ⫽ 46.3, p ⬍ .001 for group, by Fisher PLSD, p ⬍ .05 for control subjects versus depression, anxiety, and comorbid groups; depression versus
anxiety groups; and comorbid versus anxiety groups.
c
ANOVA: F ⫽ 59.3, p ⬍ .0001 for group; by Fisher PLSD, p ⬍ .002 for control groups versus depression, anxiety, and comorbid groups; depression versus
anxiety groups; and comorbid versus anxiety groups.
d
ANOVA: F ⫽ 40.9, p ⬍ .0001 for group; by Fisher PLSD, p ⬍ .05 for all group contrasts.
e
ANOVA: F ⫽ 19.7, p ⬍ .0001 for group; by Fisher PLSD, p ⬍ .05 for all group contrasts except anxiety versus comorbid groups.
f
ANOVA: F ⫽ 91.1, p ⬍ .0001 for group; by Fisher PLSD, p ⬍ .05 for all contrasts except depression versus comorbid groups.
a
b
Results
Subject Characterization
Table 1 contains age, gender, weight, diagnosis, and clinical
rating scale scores, including ANOVA and post hoc comparison
results. Matching ensured that all group comparisons included
only groups that had matched age and gender distributions.
There were no meaningful group differences in body weight.
Social phobia was the predominant anxiety disorder diagnosis.
Hamilton Depression, Hamilton Anxiety, STAI State, STAI Trait,
TSST STAI-State, and Sheehan Disability scores all differed
significantly across groups. Except for the TSST STAI-State,
ratings for the depression and comorbid groups always exceeded
the anxiety group. For all three, TSST VAS scores, anxious,
comorbid, and depressed groups differed significantly from
control subjects (range for overall F scores: 9.05–15.4, all p ⬍
.0001), but not each other for “nervous” and “anxious,” whereas
for “fearful,” the comorbid group differed from all others (which
did not differ from each other).
GH Response to Clonidine
The RM-ANOVA comparing all patients to all control subjects
(Figure 1) showed a highly significant main effect of clonidine
(i.e., repeated-measure: F ⫽ 19.11, df ⫽ 7, 651, p ⬍ .0001),
reflecting robust GH release in response to clonidine. The
clonidine-by-group interaction was significant (F ⫽ 2.48, df ⫽ 7,
651, p ⫽ .016) due to both some elevation in GH in patients at the
first two of three baseline samples and their blunted GH responses following clonidine. (Immediately before clonidine, at
9:00 AM, GH levels for patients and control subjects were
approximately equal.) The main effect of group was not significant in this or any subsequent GH RM-ANOVA comparing patient
groups to control subjects.
The three control groups for the three DSM-IV-defined diagnostic groups did not differ significantly from each other. The
RM-ANOVAs comparing each diagnostic group (pure anxiety,
pure depression, comorbid) to respective control subjects
showed the following (Figure 2): A clonidine effect on GH
release was evident in all three analyses (F ⫽ 8.2, df ⫽ 7, 196 for
anxiety, F ⫽ 3.5, df ⫽ 7, 196 for depression, F ⫽ 9.7, df ⫽ 7, 231
for comorbid; p ⬍ .0001 for anxiety and comorbid, p ⬍ .002 for
depression). The clonidine-by-group interaction was significant
Figure 1. Comparison of growth hormone (GH) levels (mean ⫾ SE) for all
anxious, depressed, and comorbid subjects combined (“patients”: n ⫽ 48) to
all individually age- and gender-matched control subjects (n ⫽ 48) before
and after intravenous clonidine (2.0 ␮g/kg) administration at 9:00 AM. x axis:
discrete time points of GH sampling; y axis: GH levels (ng/mL). See Results for
statistical analyses.
www.elsevier.com/locate/biopsych
878 BIOL PSYCHIATRY 2004;56:875– 883
O.G. Cameron et al
The pure depressed group had normal GH clonidine responses
(interaction F ⫽ .71,df ⫽ 7 196, p ⫽ .66). The comorbid group,
like pure depression, did not show significantly abnormal GH
responses (interaction F ⫽ .70, df ⫽ 7, 231, p ⫽ .67); however, in
the comorbid group GH responses appear to be intermediate,
between clear blunting seen in pure anxiety and normal responses seen in pure depression.
To further evaluate the source of blunting, patients in the
comorbid group were divided according to predominant diagnosis. All patients with predominant anxiety (pure plus comorbid) were compared with their control subjects, and all with
predominant depression (pure plus comorbid) were compared
with theirs. The RM-ANOVA for predominant anxiety showed the
robust effect of clonidine on GH (F ⫽ 11.20, df ⫽ 7, 301, p ⬍
.0001). It also showed a significant clonidine-by-group interaction (F ⫽ 3.19, df ⫽ 7, 301, p ⫽ .003), due to highly blunted GH
responding in the anxious patients (Figure 3). The parallel
Figure 2. Comparison of growth hormone levels for anxious subjects (upper
panel, “pure anxiety”: n ⫽ 15), depressed subjects (middle panel, “pure
depression”: n ⫽ 15), and subjects comorbid for anxiety and depression
(lower panel, “comorbid”: n ⫽ 18) to respective matched control subjects
(n ⫽ that for patient group for each comparison). Other details same as
Figure 1.
only in the pure anxiety group (F ⫽ 2.58, df ⫽ 7,196, p ⫽ .015),
because only the pure anxiety group showed clearly blunted GH
responses following clonidine compared with control subjects.
www.elsevier.com/locate/biopsych
Figure 3. Comparison of growth hormone levels for predominantly anxious
subjects (upper panel, “predominant anxiety patients”: n ⫽ 23) and predominantly depressed subjects (lower panel, “predominant depression patients”: n ⫽ 24) to respective matched control subjects (“matched control
subjects”: n ⫽ that for patient group for each comparison). Other details
same as Figure 1.
O.G. Cameron et al
RM-ANOVA comparing patients with predominant depression
with their control subjects showed the clonidine effect (F ⫽ 9.34,
df ⫽ 7, 322, p ⬍ .0001), but the clonidine-by-group interaction
was not significant (F ⫽ .49, df ⫽ 7, 322, p ⫽ .84) because of very
similar GH responses in depressed patients and control subjects.
Interaction terms for RM-ANCOVAs, using the 8:30 AM preclonidine GH levels as covariates (time of maximum baseline
difference across groups), remained significant for pure anxiety
and predominant anxiety analyses (F ⫽ 2.13 and 2.98, p ⫽ .042
and .005, respectively), but not for the all-patient comparison (F
⫽ 1.42, p ⫽ .19). Anxious subjects were divided into social
phobia and panic disorder subgroups. Inspection of results (not
shown) indicated comparable blunting, but subgroups were too
small to reach significance.
In summary, patients overall showed abnormal GH
clonidine responses. Blunting was clearly associated with
anxiety and not with depression. To verify results, paired t
tests compared patient groups to matched control subjects on
peak GH clonidine response (change scores). Comparison of
predominant anxiety subjects to control subjects reached
significance (t ⫽ 2.87; df ⫽ 44; p ⫽ .03), and comparison of
pure anxiety patients to control subjects showed a strong
trend (t ⫽ 1.99; df ⫽ 28; p ⫽ .06). In contrast, comparisons of
predominantly depressed (t ⫽ .55; df ⫽ 46; p ⫽ .69) and pure
depressed (t ⫽ .52; df ⫽ 28; p ⫽ .61) to their respective control
groups did not approach significance.
Because blunting of GH clonidine responses in depression
was reported in prior studies as specific to melancholia, effect of
the presence or absence of melancholia on GH response was
examined. All patients with depression were compared with their
control subjects, dividing the depressed group into those with
melancholia and those without. No significant group differences
were detected (mean ⫾ SD, ng/mL, for postclonidine peak:
melancholia ⫽ 5.37 ⫾ 4.26; nonmelancholia ⫽ 3.97 ⫾ 4.20;
control subjects ⫽ 5.27 ⫾ 4.97).
Figures 2 and 3 suggest that GH levels declined across the
baseline period in anxious patients, but considerably less so in
nonanxious patients. Exploring this further, RM-ANOVAs were
performed including the three samples before clonidine (8:30,
8:45, and 9:00 AM.). Over this baseline period, all comparisons
with control subjects involving pure anxiety subjects (groupings:
all patients; all anxiety, including all comorbid; predominant
anxiety; and pure anxiety) showed significant time effects (F
range: 4.18 –9.82; df range: 2, 56 –2, 188; all p ⱕ .02) because of
declining values in patients from 8:30 to 9:00 AM. None of the
comparable comparisons that did not include pure anxiety
subjects (groupings: pure depression; all depression, including
all comorbid; predominant depression; comorbid only) showed
similar statistically significant time effects. For the anxiety-predominant analysis, the interaction term was also significant (F ⫽
3.09; df ⫽ 4, 184; p ⬍ .02), indicating that the GH level decline
was significantly greater in predominantly anxious patients than
in their control subjects.
To determine the relationship of GH levels before and after
clonidine, regressions were performed for 8:30 AM and 9:00 AM
GH levels with peak change score and with postclonidine AUCs.
Of primary interest was the question of whether higher preclonidine levels predict reduced GH release following clonidine.
In regressions including all 96 subjects, GH levels at 8:30 and 9:00
AM were significantly predictive of postclonidine GH secretion
(AUC: R ⫽ .22, p ⫽ .03; and R ⫽ .36, p ⫽ .0004, respectively);
however, higher levels before clonidine were associated with
BIOL PSYCHIATRY 2004;56:875– 883 879
higher levels after. Baseline levels were also positively predictive
of postclonidine peak GH, although this was only significant for
9:00 AM (R ⫽ .23, p ⫽ .006).
To determine whether blunted GH responses were due to
persistently increased GH secretion, resulting in increased negative feedback, 8:30 AM somatomedin-C was measured. For pure
depression and control subjects, results (mean ⫾ SD, ng/mL)
were 234 ⫾ 24 and 221 ⫾ 12; for comorbid, 283 ⫾ 35 and 342 ⫾
42; and for pure anxiety, 292 ⫾ 24 and 273 ⫾ 21. All differences
were nonsignificant.
Regression analyses confirmed that neither age nor weight
had any impact on GH responses, using either AUC or peak
change score results. Across all 96 subjects, a large gender effect
was observed. In an RM-ANOVA with two independent variables, gender and diagnosis, the main effect of gender was highly
significant (F ⫽ 17.5, df ⫽ 1, 92, p ⬍ .0001), due to much higher
overall GH levels in female subjects; however, the clonidine-bydiagnosis interaction effect was also significant (F ⫽ 2.17, df ⫽ 7,
637, p ⫽ .03), due to reduced GH responses in both male and
female patients relative to their same-gender control subjects.
There were no other significant interactions. Thus, blunted GH
responses seen in patients were independent of gender. Additionally, ANOVAs with gender and diagnosis, using GH peak
change scores, found no gender effect, indicating that gender
differences were due to group differences in GH baselines, not
GH clonidine responses.
Regressions determined whether dimensional ratings of anxiety (STAI and Hamilton Anxiety Scores) on the same day as the
clonidine study would predict GH results (8:30 and 9:00 AM
levels, AUC, or peak change score). None significant. The pure
anxiety group had the clearest GH blunting, but also significantly
lower scores on anxiety rating scales. Examination of the relationship between depression severity (using Hamilton Depression scores) and GH measures showed no significant correlations, looking at all subjects together or patients alone (all R ⫽
.1). In contrast, TSST STAI State regressions with GH AUC
showed significant negative associations within the predominant
anxiety group (R ⫽ .47, p ⫽ .022) and within the pure plus
comorbid anxiety group (R ⫽ .42, p ⫽ .014), but not within the
normal or depressed groups.
BP and “Drowsy” Responses to Clonidine
Effects of clonidine on sedation (VAS “drowsy”) and BP
(systolic [SBP] and diastolic [DBP]) were assessed, with RMANOVAs comparable to those run on GH data. Predictably, both
SBP and DBP fell significantly over time in response to clonidine
(F ⫽ 27.7 and 48.5, df ⫽ 9, 846 for both, both ps ⬍ .0001, for
main effect of time), but patients and control subjects did not
differ in BP levels or clonidine responses. There were no
significant interactions for either SBP or DBP in comparisons of
all patients to all control subjects or any subgroup to matched
control subjects.
There was the expected rise and then fall in drowsiness in all
subjects in response to clonidine (F ⫽ 15.1, df ⫽ 6, 552, p ⬍
.0001). All patients rated themselves as more drowsy, compared
with all control subjects (F ⫽ 9.58, df ⫽ 1, 94, p ⬍ .003);
however, clonidine had similar (i.e., parallel) patterns of effect
on drowsiness ratings in patients and control subjects (no
significant interaction). There were no significant interactions for
drowsiness for any patient subgroup compared with matched
control subjects.
www.elsevier.com/locate/biopsych
880 BIOL PSYCHIATRY 2004;56:875– 883
Discussion
The primary goal of this study was to explore relative
contributions of anxiety, depression, and their comorbidity to the
blunted GH clonidine responses reported in previous research in
patients with major depression and various anxiety disorders.
Results support the hypothesis that the presence of anxiety per se
(specifically, panic disorder and social phobia, of sufficient
symptom number, severity, and chronicity to qualify for DSM-IV
diagnoses) accounts for GH blunting. In this study, pure depression was associated with a normal GH response. The comorbid
state did not affect results beyond the impact of anxiety. Thus, in
past studies of depression, failure to exclude subjects with
comorbid anxiety could have contributed to the blunting reported; however, lack of significant blunting in a sizeable group
with documented depressive–anxious comorbidity in this study
suggests that comorbidity alone probably does not fully explain
past positive results reported in the depression literature. Many
prior depression studies did not adequately control for gender,
age, weight, menstrual phase, menopausal status, and prior
recent tricyclic antidepressant exposure. This study indicates that
with appropriate control subjects (as described in Methods) and
exclusion of anxious comorbidity, depression itself is not associated with abnormal GH clonidine responses. Also, based on
rating scales, overall syndrome severity (anxiety, depression, and
disability scores on clonidine study day) was least among patient
groups for anxious subjects, indicating that acute severity per se
did not produce blunting. Furthermore, within the comorbid
group, the observed nonsignificant blunting appeared to be due
to individuals who were predominantly anxious, suggesting that
these individuals might be different pathophysiologically from
those with predominant depression.
Reports in the depression literature have suggested that GH
blunting might be linked to melancholia (Amsterdam et al 1989b;
Checkley et al 1984; Matussek et al 1980). In this study, individuals with melancholic depression showed no greater blunting
than nonmelancholic patients. Using symptom ratings, despite
several intervening days, TSST STAI State ratings for anxious
subjects were negatively correlated with GH AUC. In contrast,
anxiety ratings for the anxious subjects were lower during the
clonidine procedure, and no relationships with any GH measures
were seen. Furthermore, significant relationships were not observed for nonblunted depressed subjects with anxiety ratings,
nor for any subjects with depression ratings, during clonidine.
Thus, a measure that partly reflects depression severity—melancholia—is not associated with blunting, whereas elevated anxiety—produced here by the TSST—is negatively correlated with
GH response to clonidine, but only in those individuals with
chronic, impairing anxiety that passes the DSM-IV diagnostic
threshold.
Blunted GH clonidine responses have traditionally been
interpreted as reflecting abnormal alpha2 adrenoreceptor function (Siever et al 1981), but clonidine also binds to imidazoline
receptors, which could be implicated in GH blunting. As expected, clonidine produced significant BP decreases, but no
interaction effects were observed. Because hypotensive effects of
clonidine are thought to be mediated by imidazoline receptors,
lack of effects of diagnosis on BP response implies normal
imidazoline receptor responsivity to clonidine in all groups and
lends support to the hypothesis of abnormality of alpha2 adrenoreceptors. One prior study (Piletz et al 1996) reported normal
imidazoline binding in generalized anxiety.
Sedative effects of clonidine are believed to be mediated by
www.elsevier.com/locate/biopsych
O.G. Cameron et al
alpha2 adrenoreceptors. Drowsiness ratings demonstrated
higher levels for all disorders combined than for control
subjects. This pattern was true for individual patient groups as
well; however, differences were as large before clonidine as
after, indicating that the effect observed was not specifically
differential sensitivity to the sedative effects of clonidine.
Thus, in this study, sedative responses to clonidine do not
provide evidence for or against any abnormality of alpha2
adrenoreceptor function.
Normal somatomedin-C levels indicate that a chronic abnormality of GH axis activity, leading to abnormal feedback effects,
was not a mechanism responsible for the blunting observed. In
contrast, a significant association between GH levels and elevated anxiety symptoms, in those with anxiety disorders only,
indicates that chronic or recurrent anxiety symptoms (or both)
are probably associated with changes in either receptor function
(“subsensitivity,” “downregulation”) or other functional characteristics of the GH control system involving relevant receptors.
Separate analyses of baseline data suggested that anxious
patients, in addition to showing blunted GH clonidine responses,
might also have had higher GH levels following intravenous
catheter insertion and study initiation. Evidence for this was not
robust; patients in the pure anxiety group did not show significantly elevated GH levels relative to their matched control
subjects at any specific preclonidine time point. They did have
the highest initial preclonidine GH levels, however, and showed
declines in GH levels across the baseline period that were
considerably steeper than seen in other groups. Patients in the
predominant anxiety disorder group were the only ones who
showed significantly greater declines across the baseline period
than matched control subjects. Evidence from prior research
demonstrates that GH is acutely responsive to stress (Abplanalp
et al 1977; Brown and Heninger 1975; Kosten et al 1984; Noel et
al 1976; Rose and Hurst 1975), and GH elevations have been
observed in anxious individuals without drug administration.
Anxiety patients might have had greater GH reactivity to study
conditions (e.g., fear or novelty) than nonanxious patients.
Consistent with the hypothesis that this GH elevation is circumstance-related, resting GH levels in panic disorder are not
abnormal (Abelson et al 2005). Future research addressing this
issue is needed before drawing firm conclusions. If such hyperreactivity is present in anxiety patients, it might represent a
phenomenon (noradrenergic or nonnoradrenergic) separate
from whatever mechanism produces GH blunting, because there
was no significant relationship between baseline GH levels and
degree of blunting following clonidine.
The data showed a substantial gender effect on baseline GH
levels, with women showing higher levels than men, both before
and after clonidine. This finding supports the importance of
always using control subjects matched for gender (and other
relevant factors) in GH studies. Careful use of specifically
matched control subjects and reliance on analyses that compared
patients with their specific control group ensured that this gender
effect could not influence any of the reported diagnosis-related
effects.
One caveat in interpreting these results is the use of DSM-IV
definitions of pure depression and pure anxiety. Milder, nonexclusionary symptoms of the other pure disorder might have been
present in either or both pure diagnostic group. Our results
nonetheless indicate that the DSM-IV definitions used identified
groups different in a physiologically meaningful way. Second is
use of the categorization “predominant.” Although “principal
Axis I diagnosis” is specified as a categorization by the SCID and
O.G. Cameron et al
defined as the main focus of clinical attention or the reason for
the clinical encounter, it is not commonly used in research
studies. Thus, predominant, defined for this study as the disorder
having greater negative impact as judged by both subject and
investigators, has not yet been shown to be a reliable and valid
rating; however, as with the first caveat, it did identify groups that
differed in a meaningful way. Third, unlike some (but not all)
prior studies, no placebo control subjects were used. Fourth, this
study involved only one measure of brain noradrenergic activity.
No conclusion can be made regarding the status of brain
noradrenergic function other than that assessed by GH responses
to the alpha2 adrenoreceptor agonist clonidine, mediated by
GH-releasing hormone and somatostatin, and potentially involving other receptors and neurotransmitters (where the source of
any identified functional abnormality might actually reside) in
addition to this noradrenergic system (Frohman et al 1992;
Reichlin 1989).
These results for panic disorder and social phobia are consistent with findings from prior studies demonstrating that the
pathophysiologies of several anxiety disorders within the
DSM-IV diagnostic schema involve noradrenergic (and other)
brain circuits controlling GH release, including alpha2 adrenoreceptors. Particularly, they support prior evidence for noradrenergic abnormalities in social phobia (Gelernter et al 2004; Stein et
al 1992; Tancer et al 1993). Although the benefit of serotonergically active drugs in social phobia implies serotonin involvement
(Fedoroff et al 2001; Van Ameringen et al 2003), they do not
necessarily act at the primary site of physiologic dysfunction, and
their effectiveness is not evidence against involvement of other
systems. Further research is needed to determine which anxiety
disorders demonstrate consistent evidence for noradrenergic
dysfunction.
We hypothesized and observed that, with proper study
design, major depression would not be associated with a blunted
GH response; however, because many prior depression studies
did report such blunting, this finding requires careful replication
before full acceptance. Multiple factors in prior depression
studies probably contributed to the blunting seen. GH elevation
in anxiety disorders due to novelty or stress (or both) also
requires replication.
Future research will need to determine mechanism(s) underlying GH blunting. This study did not find evidence for abnormalities in GH axis feedback status as reflected by somatomedin-C levels, GH blunting due to elevated preclonidine GH
levels, or clonidine-related imidazoline receptor dysfunction.
Diagnostically and situationally relevant symptoms (but not
overall syndrome severity) were associated with blunting, even
though separated on average by approximately a week, suggesting that anxiety symptoms in individuals with panic disorder or
social phobia represent a correlate of tonic changes in noradrenergic function. Based on blunting in predominantly anxious
comorbid individuals, but not in predominantly depressed, the
anxious– depressive comorbid state might not be physiologically
homogeneous. Potential involvement of brain trophic factors
related to GH control, such as insulinlike growth factors I and II
(Harel and Tannenbaum 1992; Schneider et al 2003), should be
evaluated. Potential relationships of this noradrenergic system to
other systems such as the hypothalamic–pituitary–adrenocortical
system, which is known to be abnormal in depressive disorders
(Amsterdam et al 1989a; Boyer 2000; Gold and Chrousos 2002;
Kiraly et al 1997; Young et al 1994), in both healthy individuals
and those individuals with these and other (including comorbid)
disorders, also require investigation. Finally, the question of
BIOL PSYCHIATRY 2004;56:875– 883 881
whether the physiologic abnormality underlying GH blunting
reflects an effect of anxiety, an associated dysfunction, or a cause
must be addressed. Future studies can use GH clonidine responding, for example, to evaluate CNS noradrenergic tone, or as
a method of physiologic differentiation of phenomenologically
related syndromes such as in genetic studies.
This work was supported by Grant No. MH57751 (all authors), MH01931 (EAY), and MO1 RR00042 (University of
Michigan General Clinical Research Center). We thank Kathleen
Singer R.N. and the staff of the General Clinical Research Center
at the University of Michigan Medical Center for their assistance
in completing this research.
Abelson JL, Curtis GC, Uhde TW (2005): Twenty-four hour growth hormone secretion in patients with panic disorder. Psychoneuroendocrinology 30:72–79.
Abelson JL, Glitz D, Cameron OG, Lee MA, Bronzo M, Curtis GC (1991):
Blunted growth hormone response to clonidine in patients with generalized anxiety disorder. Arch Gen Psychiatry 48:157–162.
Abelson JL, Glitz D, Cameron OG, Lee MA, Bronzo M, Curtis GC (1992):
Endocrine, cardiovascular and behavioral responses to clonidine in patients with panic disorder. Biol Psychiatry 32:18 –25.
Abplanalp JM, Livingston L, Rose RM, Sandwisch D (1977): Cortisol and
growth hormone responses to psychological stress during the menstrual cycle. Psychosom Med 39:158 –177.
Altomonte L, Zoli A, Mirone L, Berchicci M, Travaglini R, Pellicano P, Greco
AV (1988): Impaired growth hormone response to clonidine in obesity. Exper Clin Endocrinol 92:231–234.
Amsterdam JD, Maislin G (1990): Comparison of growth hormone response
after clonidine and insulin hypoglycemia in affective illness. Biol Psychiatry 28:308 –314.
Amsterdam JD, Maislin G, Gold P, Winokur A (1989a): The assessment of
abnormalities in hormonal responsiveness at multiple levels of the hypothalamic-pituitary-adrenocortical axis in depressive illness. Psychoneuroendocrinology 14:43– 62.
Amsterdam JD, Maislin G, Skolnick B, Berwish N, Winokur A (1989b): Multiple
hormone responses to clonidine administration in depressed patients
and healthy volunteers. Biol Psychiatry 26:265–278.
Ansseau M, Von Frenckell R, Cerfontaine JL, Papart P, Franck G, TimsitBerthier M, et al (1988): Blunted response of growth hormone to
clonidine and apomorphine in endogenous depression. Br J Psychiatry
153:65–71.
Aston-Jones G, Rajkowski J, Cohen J (1999): Role of locus coeruleus in attention and behavioral flexibility. Biol Psychiatry 46:1309 –1320.
Balldin J, Berggren U, Eriksson E, Lindstedt G, Sundkler A (1993): Guanfacine
as an alpha-2-agonist inducer of growth hormone secretion—a comparison with clonidine. Psychoneuroendocrinology 18:45–55.
Balon R, Yeragani VK, Pohl R, Muench J, Berchou R (1990): Somatic and
psychological symptoms during isoproterenol-induced panic attacks.
Psychiatry Res 32:103–112.
Bamberger CM, Monig H, Mill G, Godde E, Schulte HM (1995): Growth hormone secretion in response to a new centrally acting antihypertensive
agent moxonidine in normal human subjects: Comparison to clonidine.
Exper Clin Endocrinol Diab 103:205–208.
Barbee JG (1998): Mixed symptoms and syndromes of anxiety and depression: Diagnostic, prognostic and etiologic issues. Ann Clin Psychiatry
10:15–29.
Bousquet P, Dontenwill M, Greney H, Feldman J (1998): I1-imidazoline receptors: An update. J Hypertension 16(suppl):S1–S5.
Boyer P (2000): Do anxiety and depression have a common pathophysiological mechanism? Acta Psychiatr Scand 406(suppl):24 –29.
Brambilla F, Perna G, Garberi A, Nobile P, Bellodi L (1995): Alpha 2-adrenergic
receptor sensitivity in panic disorder: I GH responses to GHRH and
clonidine stimulation in panic disorder. Psychoneuroendocrinology 20:
1–9.
Bremner JD, Krystal JH, Southwick SM, Charney DS (1996): Noradrenergic
mechanisms in stress and anxiety: I Preclinical studies. Synapse 23:28 –
38.
www.elsevier.com/locate/biopsych
882 BIOL PSYCHIATRY 2004;56:875– 883
Brown WA, Heninger G (1975): Cortisol, growth hormone, free fatty acids
and experimentally evoked affective arousal. Am J Psychiatry 132:1172–
1176.
Cameron OG (1994): The symptoms of adrenergic activation. In: Cameron
OG, editor. Adrenergic Dysfunction and Psychobiology. Washington, DC:
American Psychiatric Press 237–255.
Cameron OG, Lee MA, Curtis G, McCann DS (1987): Endocrine and physiological changes during “spontaneous” panic attacks. Psychoneuroendocrinology 12:321–331.
Cameron OG, Minoshima S (2002): Regional brain activation due to pharmacologically induced adrenergic interoceptive stimulation in humans.
Psychosom Med 64:851– 861.
Cameron OG, Nesse RM (1988): Systemic hormonal and physiological abnormalities in anxiety disorders. Psychoneuroendocrinology 13:287–307.
Cameron OG, Zubieta JK, Grunhaus L, Minoshima S (2000): Effects of yohimbine on cerebral blood flow, symptoms and physiological functions in
humans. Psychosom Med 62:549 –559.
Charney DS, Heninger GR, Breier A (1984): Noradrenergic function in panic
anxiety. Effects of yohimbine in healthy subjects and patients with agoraphobia and panic disorder. Arch Gen Psychiatry 41:751–763.
Charney DS, Heninger GR, Sternberg DE (1982a): Failure of chronic antidepressant treatment to alter growth hormone response to clonidine.
Psychiatry Res 7:135–138.
Charney DS, Heninger GR, Sternberg DE, Hafstad KM, Giddings S, Landis DH
(1982b): Adrenergic receptor sensitivity in depression. Effects of
clonidine in depressed patients and healthy subjects. Arch Gen Psychiatry 39:290 –294.
Charney DS, Woods SW, Krystal JH, Nagy LM, Heninger GR (1992): Noradrenergic neuronal dysregulation in panic disorder: The effects of intravenous yohimbine and clonidine in panic disorder patients. Acta Psychiatr
Scand 86:273–282.
Checkley SA, Glass IB, Thompson C, Corn T, Robinson P (1984): The GH
response to clonidine in endogenous as compared with reactive depression. Psychol Med 14:773–777.
Checkley SA, Slade AP, Shur E (1981): Growth hormone and other responses
to clonidine in patients with endogenous depression. Br J Psychiatry
138:51–55.
Coplan JD, Papp LA, Martinez J, Pine D, Rosenblum LA, Cooper T, et al (1995):
Persistence of blunted human growth hormone response to clonidine in
fluoxetine-treated patients with panic disorder. Am J Psychiatry 152:
619 – 622.
Correa H, Duval F, Claude MM, Bailey P, Tremeau F, Diep TS, et al (2001):
Noradrenergic dysfunction and antidepressant treatment response. Europ Neuropsychopharmacol 11:163–168.
Curtis GC, Nesse R, Buxton M, Lippman D (1979): Plasma growth hormone: Effect of anxiety during flooding in vivo. Am J Psychiatry 136:
410 – 414.
Fedoroff IC, Taylor S (2001): Psychological and pharmacological treatments
of social phobia: A meta-analysis. J Clin Psychopharmacol 21:311–324.
Fornai F, Blandizzi C, del Tacca M (1990): Central alpha-2 adrenoceptors
regulate central and peripheral functions. Pharmacol Res 22:541–554.
Frohman LA, Downs TR, Chomczynski P (1992): Regulation of growth hormone secretion. Front Neuroendocrinol 13:344 – 405.
Fu CH, Reed LJ, Meyer JH, Kennedy S, Houle S, Eisfeld BS, Brown GM (2001):
Noradrenergic dysfunction in the prefrontal cortex in depression: An
[15O]H2O PET study of the neuromodulatory effects of clonidine. Biol
Psychiatry 49:317–325.
Gann H, Riemann D, Stoll S, Berger M, Muller WE (1995): Growth hormone
response to growth hormone-releasing hormone and clonidine in depression. Biol Psychiatry 38:325–329.
Gelernter J, Page GP, Stein MB, Woods SW (2004): Genome-wide linkage
scan for loci predisposing to social phobia: Evidence for a chromosome
16 risk linkage. Am J Psychiatry 161:59 – 66.
Gil-Ad I, Gurewitz R, Marcovici O, Rosenfeld J, Laron Z (1984): Effect of aging
on human plasma growth hormone response to clonidine. Mech Ageing
Dev 27:97–100.
Gold PW, Chrousos GP (2002): Organization of the stress system and its
dysregulation in melancholic and atypical depression: High versus low
CRH/NE states. Molec Psychiatry 7:254 –275.
Gurguis GN, Vitton BJ, Uhde TW (1997): Behavioral, sympathetic and adrenocortical responses to yohimbine in panic disorder patients and normal controls. Psychiatry Res 71:27–39.
www.elsevier.com/locate/biopsych
O.G. Cameron et al
Harel Z, Tannenbaum GS (1992): Synergistic interactions between insulinlike growth factors-I and -II in central regulation of pulsatile growth
hormone secretion. Endocrinology 131:758 –764.
Hasey GM, Stancer HC, Warsh JJ, Persad E (1985): Neurotransmitter metabolities and endocrine responses in depression. Prog Neuropsychopharmacol Biol Psychiatry 9:613– 617.
Hollander E, DeCaria C, Nitescu A, Cooper T, Stover B, Gully R, et al (1991):
Noradrenergic function in obsessive-compulsive disorder: Behavioral
and neuroendocrine responses to clonidine and comparison to health
controls. Psychiatry Res 37:161–177.
Horton RW, Katona CL, Theodorou AE, Hale AS, Davies SL, Tunnicliffe C, et al
(1986): Platelet radioligand and neuroendocrine challenge tests in depression. Ciba Found Symp 123:84 –105.
Johnson MR, Lydiard RB (1995): The neurobiology of anxiety disorders.
Psychiatr Clin North Am 18:681–725.
Katona CL, Healy D, Paykel ES, Theodorou AE, Lawrence KM, Whitehouse A,
et al (1993): Growth hormone and physiological responses to clonidine
in depression. Psychol Med 23:57– 63.
Kessler RC, McGonagle KA, Zhao S, Nelson CB, Hughes M, Eschleman S, et al
(1994): Lifetime and 12-month prevalence of DSM-III-R psychiatric disorders in the United States. Arch Gen Psychiatry 51:8 –19.
Kessler RC, Stang P, Wittchen HU, Stein M, Walters EE (1999): Lifetime comorbidities between social phobia and mood disorders in the US national comorbidity survey. Psychol Med 29:555–567.
Kessler RC, Stang PE, Wittchen HU, Ustun TB, Roy-Byrne PP, Walters EE
(1998): Lifetime panic-depression comorbidity in the national comorbidity survey. Arch Gen Psychiatry 55:801– 808.
Kimber J, Sivenandan M, Watson L, Mathias CJ (2001): Age- and genderrelated growth hormone responses to intravenous clonidine in healthy
adults. Growth Horm Igf Res 11:128 –135.
Kiraly SJ, Ancill RJ, Dimitrova G (1997): The relationship of endogenous
cortisol to psychiatric disorder: A review. Can J Psychiatry 42:415– 420.
Kirschbaum C, Pirke KM, Hellhammer DH (1993): The “Trier Social Stress
Test”—a tool for investigating psychobiological stress responses in a
laboratory setting. Neuropsychobiology 28:76 – 81.
Koob GF (1999): Corticotropin-releasing factor, norepinephrine and stress.
Biol Psychiatry 46:1167–1180.
Kosten TR, Jacobs S, Mason J, Wahby V, Atkins S (1984): Psychological correlates of growth hormone response to stress. Psychosom Med 46:49 –58.
Lader M (1988): Beta-adrenoceptor antagonists in neuropsychiatry: An update. J Clin Psychiatry 49:213–223.
Lee MA, Cameron OG, Gurguis GN, Glitz D, Smith CB, Hariharan M, et al
(1990): Alpha 2-adrenoreceptor status in obsessive-compulsive disorder. Biol Psychiatry 27:1083–1093.
Lydiard RB (1991): Coexisting depression and anxiety: Special diagnostic
and treatment issues. J Clin Psychiatry 52(suppl):48 –54.
Matussek N, Ackenheil M, Hippius H, Muller F, Schroder HT, Schultes H
(1980): Effect of clondine on growth hormone release in psychiatric
patients and controls. Psychiatry Res 2:25–36.
Matussek N, Laakmann G (1981): Growth hormone response in patients with
depression. Acta Psychiatr Scand 290(suppl):122–126.
Mendlewicz J, Linkowski P, Kerkhofs M, Desmedt D, Golstein J, Copinschi G,
Van Cauter E (1985): Diurnal hypersecretion of growth hormone in depression. J Clin Endocrinol Metab 60:505–512.
Mitchell P, Smythe G, Parker G, Wilhelm K, Brodaty H, Hickie I (1991): Growth
hormone and other hormonal responses to clonidine in melancholic
and nonmelancholic depressed subjects and controls. Psychiatry Res
37:179 –193.
Mitchell PB, Bearn JA, Corn TH, Checkley SA (1988): Growth hormone response to clonidine after recovery in patients with endogenous depression. Br J Psychiatry 152:34 –38.
Morris P, Hopwood M, Maguire K, Norman T, Schweitzer I (2004): Blunted
growth hormone response to clonidine in post-traumatic stress disorder. Psychoneuroendocrinology 29:269 –278.
Nesse RM, Cameron OG, Curtis GC, McCann DS, Huber-Smith MJ (1984):
Adrenergic function in patients with panic anxiety. Arch Gen Psychiatry
41:771–776.
Noel GL, Dimond RC, Earll JM, Frantz AG (1976): Prolactin, thyrotropin and
growth hormone release during stress associated with parachute jumping. Aviat Space Environ Med 47:543–547.
Piletz JE, Halaris A, Nelson J, Qu Y, Bari M (1996): Platelet I1-imidazoline
binding sites are elevated in depression but not generalized anxiety.
J Psychiatr Res 30:147–168.
O.G. Cameron et al
Regunathan S, Reis DJ (1996): Imidazoline receptors and their endogenous
ligands. Annu Rev Pharmacol Toxicol 36:511–544.
Reichlin S (1989): Neuroendocrinology of the pituitary gland. Toxicol Path
17:250 –255.
Ressler KJ, Nemeroff CB (2000): Role of serotonergic and noradrenergic
systems in the pathophysiology of depression and anxiety disorders.
Depr Anx 12(suppl 1): 2–19.
Robbins TW (1984): Cortical noradrenaline, attention and arousal. Psychol
Med 14:13–21.
Rose RM, Hurst MW (1975): Plasma cortisol and growth hormone responses
to intravenous catheterization. J Human Stress 1:22–36.
Roy-Byrne PP, Stang P, Wittchen HU, Ustun B, Walters EE, Kessler RC (2000):
Lifetime panic-depression comorbidity in the national comorbidity survey. Associations with symptoms, impairment, course and help-seeking.
Br J Psychiatry 176:229 –235.
Schafer SG, Kaan EC, Christen MO, Low-Kroger A, Mest HJ, Molderings GJ
(1995): Why imidazoline receptor modulator in the treatment of hypertension? Ann N Y Acad Sci 763:659 – 672.
Schittecatte M, Ansseau M, Charles G, Machowski R, Papart P, Pichot W,
Wilmotte J (1992): Growth hormone response to clonidine in male patients with panic disorder untreated by antidepressants. Psychol Med
22:1059 –1062.
Schittecatte M, Charles G, Machowski R, Wilmotte J (1989): Tricyclic wash-out
and growth hormone response to clonidine. Br J Psychiatry 154:858 – 863.
Schneider HJ, Pagatto U, Stalla GK (2003): Central effects of the somatotropic
system. Europ J Endocrinol 149:377–392.
Shear MK (1986): Pathophysiology of panic: A review of pharmacologic
provocative tests and naturalistic monitoring data. J Clin Psychiatry
47(suppl):18 –26.
Siever LJ, Insel TR, Jimerson DC, Lake CR, Uhde TW, Aloi J, Murphy DL (1983):
Growth hormone response to clonidine in obsessive-compulsive disorder. Br J Psychiatry 142:184 –187.
Siever LJ, Insel T, Uhde T (1981): Noradrenergic challenges in the affective
disorders. J Clin Psychopharmacol 1:193–206.
Siever LJ, Trestman RL, Coccaro EF, Bernstein D, Gabriel SM, Owen K, et al
(1992): The growth hormone response to clonidine in acute and remitted depressed male patients. Neuropsychopharmacology 6:165–177.
BIOL PSYCHIATRY 2004;56:875– 883 883
Siever LJ, Uhde TW (1984): New studies and perspectives on the noradrenergic receptor system in depression: Effects of the alpha 2-adrenergic
agonist clonidine. Biol Psychiatry 19:131–156.
Stanford SC (1995): Central noradrenergic neurons and stress. Pharmacol
Therap 68:297–342.
Stein MB, Tancer ME, Uhde TW (1992): Heart rate and plasma norepinephrine responsivity to orthostatic challenge in anxiety disorders. Comparison of patients with panic disorder and social phobia and normal control subjects. Arch Gen Psychiatry 49:311–317.
Sullivan GM, Coplan JD, Kent JM, Gorman (1999): The noradrenergic system
in pathological anxiety: A focus on panic with relevance to generalized
anxiety and phobias. Biol Psychiatry 46:1205–1218.
Tanaka M, Yoshida M, Emoto H, Ishii H (2000): Noradrenaline systems in the
hypothalamus, amygdala and locus coeruleus are involved in the provocation of anxiety: Basic studies [review]. Eur J Pharmacol 405:397– 406.
Tancer ME, Mailman RB, Stein MB, Mason GA, Carson SW, Golden RN (199495): Neuroendocrine responsivity to monoaminergic system probes in
generalized social phobia. Anxiety 1:16 –23.
Tancer ME, Stein MB, Uhde TW (1990): Growth hormone (GH) response to
clonidine and growth-hormone releasing factor (GRF) in normal controls. Psychoneuroendocrinology 15:253–259.
Tancer ME, Stein MB, Uhde TW (1993): Growth hormone response to intravenous clonidine in social phobia: Comparison to patients with panic
disorder and healthy volunteers. Biol Psychiatry 34:591–595.
Tyrer P (1988): Current status of beta-blocking drugs in the treatment of
anxiety disorders. Drugs 36:773–783.
Van Ameringen M, Allgulander C, Bandelow B, Greist JH, Hollander E,
Montgomery SA, et al (2003): WCA recommendations for the longterm treatment of social phobia. CNS Spectr 8(8 suppl): 40 –52.
Woods SW, Charney DS, McPherson CA, Gradman AH, Heninger GR (1987):
Situational panic attacks. Behavioral, physiologic and biochemical characterization. Arch Gen Psychiatry 44:365–375.
Young EA, Abelson JL, Cameron OG (2004): Effect of comorbid anxiety
disorders on the hypothalamic-pituitary-adrenal axis response to a social stressor in major depression. Biol Psychiatry 56:113–120.
Young EA, Haskett RF, Grunhaus L, Pande A, Weinberg VM, Watson SJ, Akil H
(1994): Increased evening activation of the hypothalamic-pituitary-adrenal axis in depressed patients. Arch Gen Psychiatry 51:701–707.
www.elsevier.com/locate/biopsych